Hedis Measure Specification

HEDIS Measure Specification Reasoning

Overview

Structured interpretation of HEDIS quality measures: denominator/numerator logic, continuous enrollment evaluation, exclusion application, NCQA audit readiness, and Star Rating-weighted care gap prioritization. Based on NCQA HEDIS Technical Specifications (MY 2024).

Usage

• Activate when interpreting HEDIS measure denominator/numerator/exclusion logic
• Activate when evaluating continuous enrollment rules or allowable gaps
• Activate when prioritizing care gaps by Star Rating weight or SDOH barriers

Core Concepts

Response Format

Apply measure logic internally. Present the final specification, rate interpretation, or gap prioritization with justification. Do not narrate enrollment evaluation steps or exclusion logic walkthrough.

1. HEDIS Measure Structure

Every HEDIS measure follows:

Eligible Population (Denominator)
  → minus Exclusions
  → equals Eligible Denominator
  → Numerator (members who met the quality criteria)
  → Rate = Numerator / Eligible Denominator

| Component | Definition | Example (CDC — Diabetes HbA1c) | |-----------|-----------|-------------------------------| | Denominator | Members eligible based on age, diagnosis, enrollment | Age 18–75, diabetes (E11.x), continuously enrolled | | Exclusions | Members removed due to clinical exceptions | Hospice, ESRD, organ transplant | | Numerator | Members who met the quality criteria | HbA1c test performed during measurement year | | Rate | Numerator ÷ (Denominator − Exclusions) | Percentage with HbA1c testing |

Five measure types: Process (service delivered), Outcome (clinical result), Structural (system capability), Patient experience (CAHPS), Utilization (resource consumption).

2. Continuous Enrollment Rules

| Rule | Definition | |------|-----------| | Measurement year | January 1 – December 31 of reporting year | | Anchor date | Date member must be enrolled through (usually Dec 31) | | Allowable gap | ≤45 days total gap permitted | | Gap counting | Calendar days without coverage; multiple gaps summed | | Enrollment source | Medical and/or pharmacy benefit, measure-dependent |

Enrollment Evaluation Decision Tree

Is the member enrolled on the anchor date?
├── NO → Exclude from denominator
└── YES
    ├── Total gap days during measurement year?
    │   ├── ≤45 days → Continuously enrolled
    │   └── >45 days → Exclude from denominator
    └── Measure requires pharmacy benefit?
        ├── YES → Verify pharmacy enrollment separately
        └── NO → Medical enrollment sufficient

3. Exclusion Logic

| Category | Applies To | Condition | |----------|-----------|-----------| | Hospice | All measures | Hospice benefit or encounter | | Deceased | All measures | Death during measurement year | | ESRD | Diabetes, kidney | N18.6, dialysis codes | | Organ transplant | Diabetes, kidney | Z94.x | | Pregnancy | BP, diabetes | O00-O9A | | Frailty + advanced illness | Age 66+, multiple | BOTH conditions required |

Evaluation rules:

1. Apply exclusions AFTER building the full denominator
2. Check the full measurement year for exclusion events
3. Frailty + advanced illness is compound — both must be present
4. Hospice overrides all other logic
5. Document which optional exclusions are applied

4. NCQA Audit Requirements

| Source | Priority | Use For | |--------|----------|---------| | Administrative claims | Primary | Denominator, exclusions, process numerators | | Electronic clinical data (ECDS) | Primary (ECDS measures) | Lab results, vitals | | Supplemental data | Secondary | Fills claims gaps (HIE lab results) | | Medical record review | Tertiary | Validation, hybrid measures |

Common Audit Findings

| Finding | Severity | Remediation | |---------|----------|-------------| | Supplemental data without source verification | High | Implement source validation | | Enrollment gap calculation error | High | Revalidate against NCQA specs | | Incorrect age calculation | Medium | Use age as of anchor date | | Duplicate member counting | High | Deduplicate on member ID | | Stale value sets | Medium | Update code sets annually |

5. Care Gap Prioritization

Prioritization Decision Tree

Is the measure triple-weighted for Star Ratings?
├── YES → High priority baseline
│   ├── Member high-risk (Charlson ≥3 or LACE ≥10)?
│   │   ├── YES → Critical priority — immediate outreach
│   │   └── NO → High priority — standard outreach
│   └── SDOH barriers (Z-codes, high ADI)?
│       ├── YES → Assign care coordinator
│       └── NO → Automated reminder sufficient
└── NO → Standard priority
    ├── >6 months remaining in measurement year?
    │   ├── YES → Schedule in next outreach batch
    │   └── NO → Escalate if feasible
    └── Process measure (screening/test)?
        ├── YES → High closure probability — include
        └── NO → Outcome measure — coordinate with PCP

Rate Interpretation

| Rate Range | Star Level | Action | |------------|-----------|--------| | ≥90th percentile | 5-star | Maintain current programs | | 75th–89th | 4-star | Targeted improvement | | 50th–74th | 3-star | Systematic outreach needed | | 25th–49th | 2-star | Intensive intervention, root cause analysis | | 10%)

• Before submitting quality data affecting reimbursement or accreditation
• Supplemental data sources change rates by >5 percentage points

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: awslabs
• Source: awslabs/hcls-agent-skills
• License: MIT-0

Install and usage instructions live in the source repository linked above.